Micro-Scholarship, What it is, How can it help me.pdf
Five Steps for Achieving Learning Analytics Success
1. Five Steps for Achieving
(Learning) Analytics Success
Ellen D. Wagner Ph.D.
Chief Research and Strategy Officer
PAR Framework
@edwsonoma
edwsonoma@gmail.com
2. Common Definitions for Today
Data
is
informa*on,
everywhere.
It
comes
in
all
kinds
and
shapes
and
sizes.
It’s
not
all
digital,
but
most
of
it
is.
Analy(cs
are
methods
and
tools
to
parse
streams
of
digital
bits
and
bytes
into
meaningful
pa>erns
that
can
be
explored
to
help
stakeholders
make
more
effec*ve
decisions.
Learning
analy(cs
are
methods
and
tools
needed
to
parse
the
stream
of
digital
bits
into
meaningful
pa>erns
that
explore
dimensions
of
cogni*on,
instruc*on
and
academic
experience,
including
student
success.
Data-‐readiness
ranges
from
essen*al
individual
knowledge
and
skills
to
ins*tu*onal
capacity
for
crea*ng
a
culture
that
values
evidence-‐based
decision-‐making.
13. Data Readiness in Higher Ed
Analy*cs
have
ramped
up
everyone’s
expecta*ons
of
personaliza*on,
accountability
and
transparency.
Academic
enterprises
simply
cannot
live
outside
the
ins*tu*onal
focus
on
tangible,
measurable
results
driving
IT,
finance,
recruitment
and
other
mission
cri*cal
concerns.
14. While Big Data raise expectations,
student data drive big decisions in .edu
15. Costs and Completion Rates
Source:
New
York
Times;
NCES
0
10
20
30
40
50
60
70
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2-‐yr
colleges
4-‐yr
colleges
Gradua7on
rates
at
150%
of
7me
Cohort
year
20. What do we want?
The RIGHT Answers!!
When
do
we
want
them?
NOW!!
21. The Predictive Analytics Reporting
(PAR) Framework
• PAR
is
a
na*onal,
non-‐profit
mul*-‐ins*tu*onal
collabora*ve
focused
on
ins*tu*onal
effec*veness
and
student
success.
• PAR
is
a
“big
data”
analysis
effort
using
predic7ve
analy7cs
to
iden*fy
drivers
related
to
loss
and
momentum
and
to
inform
student
loss
preven7on
• PAR
member
ins*tu*ons
voluntarily
contribute
de-‐
iden7fied
student
records
to
create
a
single
federated
database.
• Descrip*ve,
inferen*al
and
predic*ve
analyses
have
been
used
to
create
benchmarks,
ins*tu*onal
predic7ve
models
and
to
map
student
success
interven7ons
to
predictor
behaviors
22.
23. Analysis/Modeling Process
• Analysis
and
model
building
is
an
itera7ve
process
• Around
70-‐80%
efforts
are
spent
on
data
explora*on
and
understanding.
24. Structured, Readily Available Data
• Common
data
defini*ons
=
reusable
predic*ve
models
and
meaningful
comparisons.
• Openly
published
via
a
cc
license
@
h>ps://
public.datacookbook.co
m/public/ins*tu*ons/
par
25. PAR Outputs
Descrip7ve
Benchmarks
Show
how
ins*tu*ons
compare
to
their
peers
in
student
outcomes,
by
scaling
a
mul7-‐
ins7tu7onal
database
for
benchmarking
and
research
purposes.
Predic7ve
Models
Iden*fy
which
students
need
assistance,
by
using
in-‐depth,
ins7tu7onal
specific
predic7ve
models.
Models
are
unique
to
the
needs
and
priori*es
of
our
member
ins*tu*ons
based
on
their
specific
data.
Ins*tu*ons
address
areas
of
weakness
iden*fied
in
benchmarks
and
models
by
scaling
and
leveraging
a
member,
data
and
literature
validated
framework
for
examining
interven*ons
within
and
across
ins*tu*ons
(SSMx)
Interven7on
Matrix
26. Faculty
Student
Success
IT
Academic
Affairs
Enrollment
Management
Financial
Aid
Ins*tu*onal
Research
PAR
is
redefining
ins*tu*onal
conversa*ons
Students
27. 5
Steps
For
Achieving
(Learning)
Analy*cs
Success
33. THANK YOU FOR YOUR
INTEREST
For
more
informa*on
about
PAR
please
visit
our
website:
h>p://parframework.org
Ellen
Wagner:
Twi>er
h>p://twi>er.com/edwsonoma
Google+
edwsonoma
On
email
edwsonoma@gmail.com